Abstract Internally administered targeted radionuclide therapy (TRT) with radio-labeled molecules that deliver cytotoxic radiation to tumor has been successfully used to treat multiple cancers. Despite promising results, there is much room to improve the durable response and survival rates achieved with TRT. TRT is ideally suited for the theranostic approach to treatment because emission imaging performed before initiating a treatment cycle can be used to predict the absorbed doses (ADs) that will be delivered. Thus, the activity needed for a therapeutic effect on tumor while keeping critical organ toxicities at an acceptable level can be planned on an individualized basis. While precise treatment planning is routinely used in external beam radiotherapy, in TRT however, treatment with fixed or weight-based activities without consideration of delivered ADs continues to be the standard of care. The main barrier to dosimetry guided personalization of TRT is the lack of dosimetry tools that are valid yet practical for the clinic environment. To improve this situation the objective is to develop, validate and bring to the clinic a platform for patient-specific dosimetry-driven treatment planning that is practical for clinical use and adaptable to various TRTs. The proposed system will integrate a toolbox for SPECT/CT imaging based voxel-level dosimetry with end-to-end testing (Aim 1), validated protocols for reducing the imaging burden associated with patient specific dosimetry (Aim 2), robust dose ? outcome models that include clinical factors and imaging biomarkers as covariates (Aim 2), and an interactive user interface that the clinician can use to plan the therapy considering dosimetric and clinical factors and the resulting efficacy/toxicity trade-off (Aim 3). The system integrates new components that will be developed exploiting recent advances such as learning-based methods for low-count SPECT reconstruction and efficient image segmentation atop our existing foundation that includes a previously developed fast Monte Carlo dosimetry code. The collaboration with an industry partner with a track record in translating innovative tools for medical image analysis will help ensure clinical translation of the system. To demonstrate the capacity of the tools developed, patient studies will focus on 177Lu DOTATATE treatment of neuroendocrine tumors. This recently approved therapy is administered in four cycles with fixed activity although there is a unique opportunity to perform SPECT imaging-based lesion/organ dosimetry after each cycle to plan the next cycle. The system can be adapted to therapies with other radionuclides and targeting agents that can benefit from SPECT/CT imaging based planning such as radioligand therapy with 177Lu PSMA for prostate cancer and emerging therapies with alpha emitters. The proposed system integrates adaptations of tools developed in the past by both teams and new tools to bring a new capacity to the end user to effectively plan TRT with all data handling conveniently performed within one platform. This will have a significant positive impact because a personalized dosimetry guided approach to TRT is likely to substantially improve efficacy while maintaining low toxicity, compared with the current arbitrary ?one dose fits all? approach.